Myths About "Safety in Numbers"

Safety in Numbers and the theory behind it is what kicked this series off. A week ago, the Minneapolis Star Tribune headline read: "As Bicycle Use Climbs, Rate of Crashes With Vehicles Falls." I've seen headlines like this often enough, but THE NEXT DAY, The New York Times claimed "Data Confirm What Cyclists Knew: San Francisco Streets Are Hazardous" where they state that cyclists are getting clipped at an ever-faster rate. Earlier, the Atlanta Journal Constitution bemoaned new cyclists filling the hospitals. Puzzled, I went to "The Atlantic" that claims cycling is increasing everywhere. As I researched even further, even the League of American Bicyclists was trumpeting the "Safety in Numbers" effect in its blog. The whole "safety in numbers" theory ties back to a paper written by Peter Jacobsen, which may be found here. While John Forester pointed out problems in the methods, and even I made a post on Cycle*Dallas to suggest people shouldn't get too enthusiastic about the whole notion, THIS POST is about the unexamined MYTH BEHIND THE MYTH ABOUT WHY "Safety in Numbers" might have validity.

You see, two paragraphs in the Minneapolis article stood out:

Why is this happening? Familiarity breeds safety, one city bike advocate believes. "People are so used to seeing bicyclists -- love them or hate -- and they don't want to hit them," said Shaun Murphy, coordinator of the city's nonmotorized transportation program. Murphy said the map of hot spots for bike-car accidents doesn't show major issues around the University of Minnesota because drivers there are so used to watching for bikes.

Do bike lanes help? Murphy also credits more bike lanes. Bike lanes can benefit pedestrians as well. When bike lanes were striped on Riverside Avenue, for example, the number of bikers riding on sidewalks dropped by 87 percent, said Hilary Reeves, a spokeswoman at Transit for Livable Communities, which manages a federal grant to promote biking and walking in the Twin Cities.

So, if I'm to believe the article (remember - most crashes are simple falls and don't involve motor vehicles at all), it turns out that if there are more bicyclists, motorists no longer want to run into them because they simply get used to seeing them. Think about THAT for a moment. Motorists don't actually actively try not to run into anything as they were all taught from the time they first entered driver's ed, but they suddenly get used to cyclists and no longer want to hit them. Kumbaya. Motorists don't hit a cyclist because that cyclists was riding against traffic in the dark with no lights or reflector; the motorist hit the cyclist because "they weren't used to watching for bikes." Forgive me if I'm a little dubious about this. All but a rare lunatic fringe of psycopaths recognize that hitting a cyclist is not a good thing and hitting one is likely to prompt awkward insurance questions if nothing else.

I suppose this popular theory about why there is safety in numbers is possible, but where I live, there really are NOT a lot of cyclists, but the motorists still avoid hitting any cyclist they see and recognize, whether or NOT they are USED to seeing them. THE NOTION THAT MOTORISTS NO LONGER HIT CYCLISTS SIMPLY DUE TO A CHANGE WITHIN EACH MOTORIST AS CYCLIST NUMBERS INCREASE IS THE REAL MYTH.

Let's assume that there IS validity to the safety in numbers theory and examine why added numbers might make things safer. The popular hypothesis is that motorists simply get better at not hitting cyclists somehow. HOWEVER, THERE ARE LOTS OF OTHER REASONS THIS MIGHT BE SO. For example, embedded in the second paragraph is one reason - a drop of 87% in bikers riding on sidewalks would certainly reduce crash rates. As another alternate hypothesis, it seems LIKELY that traffic engineers consider design elements for cyclists MORE when there are more cyclists and that THIS might also explain a drop in crash rates. Certainly, traffic engineers are quick to take credit for their efforts when auto crash rates drop. I would put this as the "when cyclists are more numerous, traffic engineers stop putting them in extreme danger as often" theory. In truth, there is probably more evidence in support of this second theory than the one that motorists suddenly get mellow and attentive. Another theory is that more cyclists leads to more cyclists seeing cyclists riding sensibly and so they get better quicker simply by "monkey see monkey do." Even THIS theory seems more likely than the "kinder gentler motorist" one.

The problem is that we don't really KNOW very well what causes collision rates to drop. All we really know for sure is that riding safer is safer. The reason is that there is little recent research done involving cyclists that isn't fatally tainted with political agendas and a desire to substantiate predetermined results. While it is easy to blame or credit motorists for crash rate changes, it is MORE LIKELY that these changes are due to changes in the way that roads are designed, and the way that cyclists use those roads.

Oh, and that graph at the top of the page? The "accident rates" were generated by a random number generator in Excel. No cyclists were harmed in its production.

In the words of John Schubert, when he rightly took me to task on the subject:

"I ride where I’m seen. The human being driving the Hummer sees me, whether I’m alone or not, on Fifth Avenue or Podunk Street. He doesn’t want a collision, so he avoids the collision. What could be simpler?"

MONDAY UPDATE: WARNING - MATH AHEAD!
Well, I started crunching some numbers from the Jacobsen paper, but along the way I found some interesting anomalies that are slowing things down. First off, I think that even the most enthusiastic proponent of the "more watchful motorist" theory will concede that watchfulness is not all that's going on. For example, I found that according to the NHTSA, alcohol was involved in 38% of all car-bike cyclist fatalities in 2008 and I'm torn about how to remove clear cases where watchfulness is a non factor from the mix. What is more perplexing is that, according to the Jacobsen paper itself, sometimes it appears to get MORE dangerous for each cyclist when there are more cyclists out. In the UK, between 1984 and 1999, a doubling of cyclists would have resulted in nearly THREE TIMES as many cycling fatalities. I'll have to ponder that number a little. If we're to accept the "motorists are changing" theory, in the UK, it bodes poorly for cyclists if more ride. Regardless of anything else, I hope we can all agree that making it easier for motorists to notice us and what we are planning to do next is a sound policy. That is as true for a clueless motorist on a cell phone as for one that's almost paralyzed from fear he/she might hit a cyclist. While a motorist might well be looking for a cyclist darting off a sidewalk against traffic in the dark, any motorist being so watchful is going above and beyond the call of duty, and any cyclist that counts on such a response from a motorist is likely to fare poorly. Even in Holland.